The present invention relates to a control apparatus and a method for controlling a system equipped with a plurality of subsystems and a supervising subsystem for supervising these subsystems. More specifically, the present invention is directed to a control method and a control system capable of improving a control performance of an overall system by utilizing self condition amounts detected by the respective subsystems.
In the conventional control system equipped with a plurality of subsystems and a supervising subsystem for supervising these subsystems, as described in JP-A-2-170202, the supervising subsystem divides the supervising goal into subgoals with respect to the respective subsystems, and a subsystem which has received this subgoal evaluates its own subgoal achievement performance using its own knowledge. Further, the supervising subsystem which has received this subgoal achievement performance, again sets the subgoal, if required, to repeat the above-described process, and otherwise sends out an execution instruction to the subsystem, so that the effective control can be realized.
However, in the above-described prior art control system, when the subsystems evaluate the subgoal achievement performances, although the subsystems consider their own local knowledge, these subsystems never consider their own conditions.
Also, as one example of the conventional distributed type control system, there is described such a method of "Autonomous Distributed Type Overload Preventing Monitoring Control Logic" written by Hirata et al., Institute of Electrical Engineers of Japan, Research report RE 89-79, Jul. 20, 1989. In this conventional control system, under an initial condition such that the power systems in a plurality of regions (subsystems) are connected with each other via the tielines, one method has been proposed, which determines the power generating amounts of the respective regions in order to optimize economical subjects within such a range where the tielines are brought into an overload. In accordance with this method, since the subsystem need not consider interference with other subsystems, the respective subsystems merely execute the processes previously given to their own subsystems, so that the above-described goals can be achieved.
However, in accordance with the above-described distributed type control system, the application range of this control method is limited to the above-explained purpose, and no care is taken to such an extraordinary case of the subsystem.